Colorectal cancer (CRC), the second most commonly diagnosed cancer, is a biologically heterogeneous disease. Molecular characterization of tumors, including somatic mutations in BRAF and KRAS, microsatellite instability (MSI), and CpG island methylator phenotype (CIMP), has provided evidence of multiple tumor subtypes that develop through activation of diverse neoplastic pathways. More recently, new technologies, such as next-generation sequencing as applied in The Cancer Genome Atlas (TCGA) Project1, have enabled further characterization by identifying mutated genes in colorectal tumors, including well known genes, such as APC, SMAD4, and PIK3CA as well as some that are less well known, such as SOX9 or ACVR1B. These discoveries highlight the importance of a number of key pathways, including MAPK, Wnt, or TGF? signaling pathways. These detailed molecular data now allow us to better define tumor subtypes by mutated genes and pathways. Knowledge of the relationship between etiologic factors for CRC-such as germline genetic, lifestyle, and environmental risk factors- and tumor subtype is critical to improve our understanding of the underlying carcinogenic mechanisms that result in different molecular subtypes of CRC; however, these relations have not been comprehensively studied. To address this, we propose to use existing resources of the Genetics and Epidemiology of Colorectal Cancer Consortium (GECCO) and the Colon Cancer Family Registry (CCFR). These consortia are based on well-characterized cohort and population-based case-control studies and have available extensive germline genetic, clinical, and epidemiologic data. In addition, a large subset of cases has already undergone molecular characterization for BRAF and KRAS mutations, MSI, and CIMP. We will harmonize these existing tumor characteristics in Aim 1 in up to 11,900 CRC cases to investigate associations between defined CRC subtypes and a) common and rare germline genetic variants across the genome (Aim 1a), as well as b) lifestyle and environmental risk factors, such as alcohol, smoking, obesity, hormone use, or dietary factors (Aim 1b). In Aim 2 we will use the tumor DNA of 4,200 CRC cases by deeply sequencing approximately 100 genes identified through existing and ongoing CRC tumor tissue sequencing studies (e.g., TCGA) to define new CRC subtypes based on mutated pathways, such as Wnt, PI3K, or TGF? signaling. We will investigate associations between these subtypes and common and rare germline genetic variants across the genome (Aim 2a), as well as lifestyle and environmental risk factors (Aim 2b). This integrated approach addresses important gaps in knowledge about the risk-factor profile of existing and newly-defined molecular subtypes of CRC. Our large and well characterized study population provides a unique opportunity to define new molecular classifications with improved precision. This precision is needed to better understand how genetic and environmental risk factors, together, contribute to individual risk of CRC. These insights on carcinogenic mechanisms can help optimize prevention measures and will have an important public health impact.